Last year I developed a website called http://www.retireonproperty.com to share my own journey in property investing and share all the things I have picked up and learned along the way. With a relocation from Asia to the Middle East I have not had much chance to work on the site in recent months but am planning to pick it up again and wanted to share the following article with you and invite your comments / feedback:
The page is about historical house prices since 1960 and I have tried to present the data in an objective manner so people can make their own conclusions. I need to update the page to include data up to 2009/2010 but was hoping to do that whilst also incorporating any comments / suggestions.
If you have any suggestions on what else you’d like to see in the website I would like to hear those as I intend to continue developing it over the years and keep all the info for free.
Unmester, thanks. Will try to include a graph of house price growth versus IR, but will probably put that in a sepratae article to keep this one focused just on prices. I did make one these graphs in the past I didn’t get too much out of them after than the timing / cycle of interest up, growth down and vice versa. One issue I had was aligning the scales as the interest data I have is on a monthly basis and the growth data I calculated myself and is on an annual basis. If you don’t align the scales well then you can’t see the cycle…
Do you know a source of historical median income data as I haven’t found that yet?
I guess your logarithmic graphs represent rate of change and, like you have noticed, it is here that IR changes really start to correlate. IRs up, prices down and visa-verca.
Just work out an averge yearly IR rate from the data you have, its all you really can do to bring the 2 graphs in line.
A logarithmic vertical axis can make the data set look a lot more flat than it should do, though. Or soften the long term rate of change trend. You should be able to see the effects of this in your extrapolation. Do you really think the country can sustain an average house price of 8mil in 20 years time?
Median house price of 8 million in 20 yrs? I don't know, we might be surprised by how high it will be, but I do believe the property markets will sustain their growth in the long term but there will be some potentially significant shifts in our behavior/preferences which will reduce the median price or at least slow the growth. One thing I am sure of is that we’ll see much more medium and high density housing appear just like it has done in all of Europe. The generational changes seem to support that and I think we’ll see much more cleverly designed units where spaces can double up in functionality allowing the overall size to come down. I also expect that the concept of building townhouses around shared recreational areas (e.g. pool & gardens) will become very common.
To bring some of those stats into perspective, population growth (coupled with average number of people per household) during the period of analysis will show the growth in demand, a state by state median price (ACT/NT/WA/SA/Qld will have lagged wildly behind NSW & Vic for much of the 1960's – 1990's), have you taken into account that up until the mid 2000's stats were mainly reported as average prices not median prices?
I'm sure they were using the median price in the 90's but I could be wrong.
When you say 'average' house price can you clarify? Statistically an 'average' can be the median, the mean (several types), or the mode. Each has it's own uses. In the 2000's did they actually change the calculations or did they just clarify what type of average they were using by calling it the median price?
Go as far as you can see and you will see further.
generally, the mean is the average ie total value of sales/number of properties sold (otherwise known as the arithmetic mean, there is also a geometric mean). The median is more of a statistician’s average ie if all of the sales were listed in order of value then the list divided in two and the middle number (sale) being the median. Both methods are open to criticism, the first can be skewed by a large number of higher or lower values & the latter by the greater number of higher priced sales in the major capitals compared to total sales in regional areas.
Likewise, the stats can be skewed if the data is corrupt eg mixed data which doesn’t exclude say 2 & 4 bedroom houses as opposed to including all sales.
I have scanned through Stapledon's report and the data (will read it later in detail), but for me there are a few main points which stand out:
(1) between 1880 and 1950 there is very little growth in house prices, annualized over the period it is 1.2% (till 1949) or 2.5% (till 1950); 1950 seems to be a discontinuity in his data.
(2) between 1950 and 2006 the annualized growth is about 8.2% and in real terms (2005) that works out to less than 2.5%
If you plot the data on a log scale you can easily see points (1) and (2)
(3) Stapledon has no real explanation for this shift in growth post 1950, but to my surprise he does not touch on the issues of demand and availability of finance in respect to this structural change
(4) Stepledon also demonstrates that their is no 'natural' housing affordability measure in terms of a 'normal' price/income ratio. I will write a separate article on that in due course as it intrigues me especially with the frequent statements from all the various international bodies and e.g. the demographia report…